A chilled pack carries a use by date and a batch code printed in the same pass, and the batch code maps to a factory, a line, a shift and a day. This one says a date no chilled product gets and a batch the manufacturer says does not exist.
I am a section manager on chilled and the date checking is mine.
How the dates work, because this is the most ordinary job in the shop.
Every chilled pack has a use by date, which is a safety date rather than a quality one, and a batch or lot code printed alongside it. Both go on in the same print pass at the factory, on the same printer, usually in the same font on the film or on a label.
The code maps to a record. Factory, line, shift, date and time, and a manufacturer can tell you which tanker of milk went into it. That is the whole of traceability law and it is how a recall works.
My job twice a day is to walk the cabinets with a handheld and pull anything short dated, and the handheld knows the expected date range for every line, because it has the shelf life in it.
So nobody has to form a view about a date. The machine beeps.
In March I pulled a pack of cooked chicken slices with a use by date fourteen months away.
Not a smudge, not a misread. The date was printed cleanly, in the right place, in the right format, in the manufacturer's own print style, and it read a day in the May of the following year.
A cooked meat slice has a shelf life of about ten days.
What I did with it, which is the procedure and nothing clever.
The handheld rejected it, which is what it is for, and the procedure for a date anomaly is to quarantine the pack, record it, and raise it with the supplier through the technical team, because a mis-dated pack is a food safety incident whoever caused it.
So the pack went into a sealed bag in the quarantine fridge with a label and a number, and I filled the form in and sent photographs.
What the manufacturer came back with, eleven days later, in writing.
The batch code on that pack does not correspond to any production run at that site.
Their letter is two paragraphs. It says the code format is correct for their site and their line numbering, that the line number in it is a line they operate, and that no production is recorded under that code on any date, and that they are unable to account for the pack.
They asked for it back. The technical team couriered it to them.
The second letter, which is the one I have kept a copy of.
They examined the pack. The film is their film, from their supplier, with their print. The seal is their seal, made on their equipment, and they say so. The product inside was analysed and is consistent with their product and within specification for everything they tested.
They can tell all of that because packaging film is bought in reels with its own codes and the seal pattern on a tray is as individual as a tyre tread.
So it is their pack, made on their equipment, with a date no pack of theirs has ever carried and a batch number that is not in their records.
Their conclusion is that the pack was not produced by them and that they are reporting it internally as a potential counterfeiting matter.
What I found when I went back through my own records.
There had been others. This was the first one I noticed.
We keep the date check sheets for three months and the quarantine log for a year. In the log there are five previous entries in my own hand and two in a colleague's, over about two years, all described as date print error, all from chilled, all quarantined and binned without going to a supplier, because a date print error is a routine thing and you do not write to a factory about one pack.
I can remember three of them. I remember that they were long dates rather than short ones and I remember that I did not think about it, because a printer that prints the wrong date prints nonsense, and nonsense is what you expect.
None of them was nonsense. The log records the dates and they are all real dates, all between eleven and sixteen months out, all in the correct format.
What happens now.
Nothing, formally. The supplier's report went to their own quality system and the store had no action to take because the product never reached a customer.
Our technical team's closing note says the incident is resolved, the pack was withdrawn and the supplier has investigated.
What I do differently.
The date check is twice a day, morning and evening, and I do it the same way I always have.
What I have added, off my own bat, is that I now photograph every long dated pack before it goes in the bag, with the code legible, and I keep the photographs on my own phone.
I have eleven photographs since March.
The thing that keeps me doing it is small and I would rather it were not there. The codes in those eleven have a sequence in them, the last three digits, and across the eleven those three digits are in order, ascending, with gaps.
A batch code is sequential because a factory counts its runs. An ascending sequence with gaps, from a batch record that does not exist, means somebody or something has made at least as many runs as the highest number, and I have seen eleven of them, and the latest gap is forty one.